Building product at Stripe: craft, metrics, and customer obsession | Jeff Weinstein (Product lead)

Jul 11, 2024 Episode Page ↗
Overview

This episode features Jeff Weinstein, former Product Lead for Stripe's payment infrastructure and Atlas, discussing his "go, go, go plus optimism, long-term compounding" philosophy. He shares insights on customer communication, picking metrics, achieving product-market fit, and the Stripe Study Groups program for building beloved products.

At a Glance
23 Insights
2h 35m Duration
16 Topics
7 Concepts

Deep Dive Analysis

Jeff Weinstein's Background and Mindset

The 'Go, Go, Go ASAP + Optimistic, Long-Term Compounding' Approach

The Importance of Craft and Quality in Product Development

Effective Customer Communication and Feedback Strategies

Prioritizing Feedback from Paying Customers

The Role of Metrics in Product Success and Decision Making

Creating a 'Users Having a Bad Day' Metric

Stripe's 'Study Groups' for Product Improvement

Building Products with Customer Empathy and Unnatural Counterbalances

Stripe Atlas: Simplifying Company Formation Globally

Automating Complex Administrative Processes with Atlas

Impact of Atlas on Entrepreneurship and Economic Growth

Operational Efficiency and Automation Philosophy at Atlas

Building Diverse Teams for Better Product Outcomes

Lessons on Building New Products within a Large Company

Navigating Competition and Fostering Partnerships

Go, Go, Go ASAP + Optimistic, Long-Term Compounding

This philosophy combines immediate, energetic action with a strategic, patient approach to building foundational capabilities. It emphasizes injecting energy to get things done quickly while also investing in long-term infrastructure and services that will compound over time, such as improving API latency or reliability.

Product Market Fit Signal

A key indicator of product-market fit is when customers are so reliant on your product that they are furious or clamoring for it when it's unavailable or broken. If customers are merely murmuring or mildly inconvenienced, it suggests the product isn't solving a burning enough problem.

The Power of Silence in Customer Conversations

Instead of pitching a product, product builders should ask open-ended questions about a customer's biggest problems or frustrations and then listen without interruption. This silence allows customers to reveal their most burning needs, which can directly inform the product roadmap and uncover adjacent opportunities.

Users Having a Bad Day Metric

This is a quantitative measure where a log line is emitted every time a user encounters a problem (e.g., a 404 error, delayed payout, multiple payment declines). By tracking the frequency of these 'bad day' reasons in a stacked bar chart, teams can identify and prioritize which issues to address, even if they weren't initially aware of them.

Study Group

An internal program at Stripe where random groups of 4-8 employees (from any department) pretend to be a customer with a specific goal. Participants are forbidden from using internal Stripe knowledge and are not there to solve problems or critique, but purely to practice empathy and experience the product from an external perspective, uncovering friction points.

Proof of Existence

This concept emphasizes demonstrating tangible progress by getting a first version of something working, even if it's janky or minimal. It's a powerful way to cut through debate and theory, providing concrete evidence that a solution is possible and building momentum for further development.

Counter Positioning

A business strategy where a company adopts a new, superior business model that its competitors cannot imitate without undermining their existing successful business. This creates a competitive advantage by making it costly or self-destructive for rivals to follow suit.

?
Why is customer obsession and product craft so important for product success?

Customer obsession and craft are crucial because they ensure the product solves a truly burning problem for users, leading to strong product-market fit. Without solving a fundamental need, no amount of craft or delight will make a product successful or sustainable.

?
How can product builders effectively listen to customers without pitching their own solutions?

Product builders should ask open-ended questions like 'What's in your email?' or 'What grinded your gears last week?' and then practice silence. This creates space for customers to share their most pressing problems, allowing the product builder to identify real needs rather than guessing.

?
How can product teams narrow down customer feedback when there's an overwhelming amount?

To narrow down feedback, product teams can ask for specific details like screenshots or 3-5 bullet points via email, which helps self-select highly engaged customers. They can also create bounded programs, like a 'bug finder program' with a limited timeframe, to manage the influx of feedback.

?
Why should product teams prioritize feedback only from paying customers?

Feedback from paying customers is more reliable because they are willing to exchange money for a solution, indicating a genuine and urgent need. Friends or beta testers might give positive feedback without truly needing the product, leading to wasted development effort on non-essential features.

?
How can a single metric empower a product team and drive impact?

A single, customer-centric metric (like 'companies with zero support tickets') provides a clear, shared goal that everyone on the team can understand and rally around. It forces trade-offs, guides daily decisions, and allows engineers to identify and solve problems directly, leading to significant product improvements.

?
How can product teams improve their understanding of user problems at scale?

Teams can implement a 'users having a bad day' metric by emitting a log line for every identified negative user experience. This creates a continuous, objective counting system for problems, helping teams track frequencies, prioritize issues, and identify previously unknown pain points.

?
What is Stripe Atlas and what problem does it solve?

Stripe Atlas is a service that radically simplifies the process of starting a company, particularly for international founders who historically had to travel to the US. It automates legal, tax, and banking setup, allowing entrepreneurs to incorporate a US company in a single day, removing administrative complexity so they can focus on building their business.

?
How can new products be successfully built and launched within a large company?

Success requires aligning teams with deep customer stories, storyboarding an unconstrained vision, showing tangible forward progress quickly, and building momentum. It also involves making the product economically viable and fostering trust by transparently sharing metrics and inviting early customer involvement.

1. Prioritize Burning Customer Problems

Focus product development on problems that people desperately need solved, those that would make them “pause their entire day” or “leap through the computer” for a solution, as craft and beauty won’t compensate for a lack of core need.

2. Practice Silence in Customer Conversations

When talking to customers, avoid pitching your product; instead, sit in silence and ask open-ended questions (e.g., “What grinded your gears last week?”, “Magic wand, what do you wish you could have off your plate?”) to uncover their most burning, unarticulated problems and build your roadmap.

3. Require Payment as Forcing Function

To validate true customer need and value, ask potential customers to pay for a solution (even a small amount like $1) or commit financially, as “ready to pay” is significantly different from actual payment and reveals true priorities.

4. Focus on Go-Go-Go with Optimism

Inject immediate energy and optimism into taking action, aiming to make things “due tomorrow” rather than later, as this ignites interest and produces surprising results.

5. Balance with Long-Term Compounding

Pair the “go, go, go” attitude with a strategic, long-term compounding mindset, investing in foundational capabilities (e.g., faster APIs, reliability) that you will “never regret spending time in” and that build layers of infrastructure over time.

6. Pick Metrics Reflecting Customer Value

Select a small number of metrics that are numerical representations of the value provided to the customer, measured from their perspective, to align teams and force trade-offs.

7. Measure Zero Support Tickets

For products like Atlas, track the percentage of customers who complete the entire process (e.g., company incorporation) without needing to contact support, as this metric directly correlates with customer satisfaction and willingness to recommend.

8. Create ‘Bad Day’ Chart

Implement a system to log an event whenever a user encounters a problem (e.g., 404 error, delayed payout, multiple payment declines) and visualize these “bad day reasons” in a bar chart to identify and prioritize issues.

9. Embody Customer with Study Groups

Organize internal “study groups” where 4-8 employees (from any department) pretend to be a customer with a specific problem, strictly avoiding internal company knowledge, to foster empathy and uncover product friction points.

10. Respond to Customer Feedback Immediately

When a customer takes the time to communicate a problem, treat it as a P0 alert and respond quickly, even if it’s just to acknowledge receipt, to build trust and gather high-signal information.

11. Discount Feedback from Friends

When seeking product feedback, strictly ignore input from friends and focus solely on target customers who are willing to pay, as friendly feedback can be misleading about actual market need.

12. Prioritize Problems One and Two

As a leader, focus your energy and the team’s attention on solving the most critical, hardest problems (problems one and two), rather than getting caught up in numerous smaller, easier issues (problems three through a hundred).

13. Storyboard Unconstrained Solutions

When envisioning solutions, start by drawing the “unconstrained perfect solution” with a Sharpie (like a Pixar storyboard), rather than immediately jumping to high-fidelity designs, to foster bold thinking.

14. Seek Proof of Existence

Instead of relying on theoretical arguments or debates, demonstrate the feasibility of a new idea by achieving “proof of existence” (e.g., successfully sending one piece of mail for the 83B election) to build momentum and trust.

15. Integrate Early Customers into Meetings

Invite early customers directly into team meetings and automatically pipe their feedback into internal channels (e.g., Slack) to create constant engagement and motivate the team with direct customer stories.

16. Empower Customers to Design

For certain customer segments (e.g., founders with product skills), invite them to actively design and draw their ideal product interfaces or dashboards using collaborative tools, rather than guessing their needs.

17. Begin Sentences with the Customer

Adopt the mindset of starting every thought or discussion about product decisions by physically or mentally placing the customer first (e.g., “The customer is sitting here…”), to ensure customer-centricity.

18. Build Diverse Teams Intentionally

When building a team, be intentional about creating diversity in perspectives and backgrounds from the outset, ensuring the candidate pool matches the desired team composition, as this leads to more effective outcomes.

19. Embrace ‘Make Some Mistakes’

Encourage creativity and experimentation by explicitly stating “let’s make some mistakes” during brainstorming or early-stage development, fostering an environment free of pretense and evaluation.

20. Craft Motivating Metric Titles

Name metrics concisely and clearly (e.g., “Companies with Zero Support”) to make them feel impactful and customer-centric, fostering internal buy-in and reducing the need for constant reminders of their importance.

21. Maintain Dashboard Hygiene

Ensure dashboards are aesthetically pleasing, with consistent axes, relevant decimal places, and discoverable URLs (e.g., go/metrics), to encourage frequent team engagement and trust in the data.

22. Make Products Economically Viable

Ensure that all products, even those focused on customer acquisition or ecosystem growth, have a clear economic viability strategy and metrics to demonstrate their value and justify long-term investment.

23. Treat Competitors as Alternatives

Maintain open relationships and even shared communication channels with “competitors,” viewing them as alternatives, to foster mutual benefit, share insights (e.g., government delays), and potentially collaborate for broader mission success.

The moment the customer felt compelled enough to go out of their way to talk about some problem, that's an unbelievable gift. I will leave a meeting to just get one message back to them.

Jeff Weinstein

People don't really get out of bed for their second problem, right? They get out of bed for their first problem.

Jeff Weinstein

If you're text message friendly with five or 10 of those, you are going to have so much direct signal that is infectious.

Jeff Weinstein

If you by accidentally leaked your dashboard to them, your customer would be ecstatic to learn that that's what you were measuring the whole time.

Jeff Weinstein

You are one of the best people I've ever worked with at solving problems three through a hundred, but I need you stuck on problems one and two.

John Collison

You can't screw up a sentence if it begins with the customer.

Jeff Weinstein's Dad

If it's not on Go Metrics, I'm not going to look at it.

Jeff Weinstein

Stripe Study Group

Jeff Weinstein
  1. Gather 4-8 random people from different teams within the company.
  2. Pretend to be a fictional company (e.g., Dolphin Aquarium Industries) with a specific outcome problem (e.g., accepting in-person payments at a farmers market).
  3. Adopt roles within the fictional company (e.g., CEO, designer).
  4. Go through a product experience end-to-end, embodying the customer and not breaking character.
  5. Strictly adhere to Rule 1: You do not work at Stripe (no internal Stripe knowledge allowed).
  6. Strictly adhere to Rule 2: You are not here to solve problems or critique (focus purely on empathy and experiencing the product).
  7. Record observations and insights, which are later funneled into existing formal bug/priority processes.

Identifying Burning Customer Problems (using Silence)

Jeff Weinstein
  1. Initiate a conversation with a customer or prospective customer.
  2. Avoid pitching your own product or solutions.
  3. Ask open-ended questions about their daily work, frustrations, or desires (e.g., 'What's in your email?', 'What grinded your gears last week?', 'Magic wand, what do you wish you could have off your plate?').
  4. Practice silence after asking the question, allowing the customer to speak freely and reveal their most pressing problems.
  5. Listen for patterns in problems across multiple customer conversations to identify high-impact opportunities.
One in six
Percentage of new Delaware corporations started on Stripe Atlas As of the podcast launch date.
15%
Percentage of founders getting through Atlas with zero support tickets (initial) When Jeff Weinstein started working on Atlas.
85%
Percentage of founders getting through Atlas with zero support tickets (after 18 months) After focusing on the 'zero support tickets' metric.
More than 25
Number of Study Groups run at Stripe in early 2024 In the last few months of 2024.
More than 250
Number of people who participated in a Study Group at Stripe in early 2024 In the last few months of 2024.
Above 80
Stripe Atlas NPS score With almost a 50% response rate, considered quite high.
Low 60s
Apple's NPS score For comparison to Atlas NPS.
75
AirPods NPS score For comparison to Atlas NPS.
55,000
Number of Atlas companies to date Total companies started through Atlas.
$5 billion
Annual revenue generated by Atlas companies Per year, representing the collective GDP of Atlas companies.
More than 20%
Percentage of multi-founder teams with at least one cross-border founder Among Atlas users.
10,000
Number of 83B elections filed by Atlas Crossed by the time the podcast airs, all 100% on time.